Poster 322: Patient Acceptable Symptom State (PASS) 2 Years After Anterior Cruciate Ligament Reconstruction Reflects Knee Loading Patterns Associated with Worse Outcomes

Abstract

Objectives:

Anterior cruciate ligament (ACL) tears are common knee injuries that greatly increase osteoarthritis (OA) risk. By 1 year after ACL reconstruction (ACLR), most patients have been released to unrestricted activities. However, ACLR does not fully restore knee mechanics and altered knee loading has been associated with greater OA risk. Knee mechanics can be challenging to measure. Consequently, clinical outcomes are more commonly assessed using patient reported outcomes (PRO) such as the Knee Injury and Osteoarthritis Outcome Score (KOOS). Patient acceptable symptom state (PASS) criteria were developed to aid in the interpretation of the KOOS. This study was performed to test the following hypotheses: (1) that loading patterns of ACLR knees 2 years after ACLR differ between patients who achieve an acceptable symptom state (PASS-Y) and those who do not (PASS-N); and (2) that those who fail to reach PASS exhibit loading patterns associated with greater pain and higher OA risk.

Methods:

This IRB-approved study included 59 individuals who completed the KOOS and underwent gait analysis 2 years after primary unilateral ACLR (Table 1). The gait test acquired three walking trials at a self-selected comfortable pace using a standard 3D motion capture system (Qualisys & Bertec synchronized at 120 Hz) with the point cluster technique. External knee moments relative to the tibial frame were calculated using an inverse dynamic approach and normalized to percent body weight multiplied by height (%BW×Ht). The total joint moment (TJM) was calculated for each time point of the stance phase as the Euclidean norm of the flexion (KFM), adduction (KAM), and rotation (KRM) moments. Absolute loading metrics, peak KFM, KAM, KRM, and TJM, were extracted from early stance. Relative loading metrics were the %KFM, %KAM, and %KRM contribution to peak TJM. Participants were dichotomized based on ACLR-specific KOOS PASS-Y criteria: 88.9 (pain), 57.1 (symptoms), 100 (function in activities of daily living, ADL), 75 (function in sports and recreation, sport/rec), 62.5 (knee-related quality of life, QoL). Participant factors, surgical factors, and loading metrics between PASS-Y and PASS-N were compared using independent t-tests (or Mann-Whitney U tests if non-normally distributed; Fisher’s exact test for categorical variables). Relationships between the loading metrics and PASS were examined with mixed effects logistic regression and odds ratios were computed. Results are reported as mean (95% confidence interval (CI)).

Results:

While all participants (59/59) met PASS-Y for symptoms, only 36% (21/59) achieved PASS in all domains (Figure 1). Those meeting PASS-Y were sports/rec=43 (73%), QoL=39 (66%), pain=39 (66%), and ADL=25 (42%). Participant & Surgical Factors: Those with PASS-N for pain and ADL showed a more varus mechanical axis (-1.0 [-2.1, 0.2]°, p = 0.043; -1.3 [-2.3, -0.3]°, p = 0.014, respectively). PASS-N for ADL also showed a lower Tegner score than PASS-Y (-1.1 [-2.0, -0.0], p = 0.043). No differences in graft type or meniscus status were observed. Loading Metrics: Those with PASS-N for pain and QoL showed a lower KFM (-0.85 [-1.49, -0.22] % BW×Ht, p = 0.016; -0.75 [-0.14, -0.10] % BW×Ht, p = 0.024 respectively). Those with PASS-N for ADL had higher KAM (0.42 [0.11, 0.74] % BW×Ht, p = 0.009). No differences were observed between PASS-Y and PASS-N for KRM and TJM. Across the KOOS pain, ADL, sport/rec, and QoL subscales, PASS-N consistently demonstrated a lower %KFM and a higher %KAM compared to PASS-Y (Figure 2). Loading vs. PASS: Participants with a lower KFM had odds of 1.94 [1.11, 3.40] (p = 0.020) and 1.75 [1.04, 2.96] (p = 0.035) for not attaining PASS in pain and QoL, respectively. Participants with a higher KAM had odds of 3.53 [1.29, 9.61] (p = 0.014) for not attaining PASS in ADL. These results still applied with the addition of independent explanatory variables (i.e., mechanical axis for KFM vs. pain; Tegner for KAM vs. ADL).

Conclusions:

Nearly two-thirds of patients 2 years after ACLR did not achieve PASS in at least one KOOS subscore. Failure to achieve PASS correlated with quantitative gait data reflecting altered loading known to reduce longer term outcomes. These data demonstrated utility and sensitivity of the KOOS subscore PASS criteria for personalized assessment of outcomes early after ACLR where the relative lack of symptoms may mask suboptimal recovery. Furthermore, participants 2 years after ACLR who did not consistently reach PASS exhibited loading patterns associated with greater pain and greater OA risk than those achieving PASS-Y. Cross-sectionally greater KAM 2 years post-ACLR has been shown to correlate with medial knee degenerative changes at the same time point and to predict worse outcomes 8 years after ACLR. ACLR knees with a lower KFM than contralateral knees also exhibited altered extensor/flexor co-activation patterns. The KFM and TJM are responsive to changes in pain and are indicative of knee muscle strength and activation. This means that interventions aimed to increase KFM and to restore the relative loading contributions (%KFM & %KAM) to TJM may lead to improved quadriceps strength and pain relief. The absolute loading metrics, as suggested by the regression analyses, may also have predictive value for identifying participants who may benefit from additional interventions.

Significance:

These data suggest that KOOS PASS criteria may be an accessible screening tool to provide early warning of patients at higher risk for poor outcomes who may potentially benefit from further assessment, additional physical therapy, or closer follow-up.